Introduction

The holopelagic species of the brown alga Sargassum (S. fluitans and S. natans; sargassum from hereon) play a vital role in ocean scale processes and provision of biodiversity of the Atlantic Ocean, and their drifting masses (rafts) are also known as the “golden rainforest” (Laffoley et al. 2011). Historically, sargassum rafts have been concentrated by a subtropical Atlantic gyre in the Sargasso Sea, with periodical releases of drifting masses into the northern Caribbean Sea and the Gulf of Mexico (Acton et al. 2019; Frazier 2014).

The Sargasso Sea is mainly oligotrophic; however, combined oceanographic features, such as mesoscale eddies, upwelling gyres and winter storms, as well as biological processes, such as N fixation and recycling through microbial loop, can define localized high primary productivity (Lafolley et al. 2011; Lomas et al. 2013; Nelson and Carlson 2012). Satellite observations from 2002 to 2008, estimated a biomass of ~ 2 million wet tons in the Sargasso Sea and Gulf of Mexico (Gower and King 2011). Mixing of surface waters in the northern Sargasso Sea during the winter months, sink the sargassum and other primary producers to deeper waters (Baker et al. 2018). In addition, a considerable proportion of the primary production of the algae is released as dissolved organic carbon, a major oceanic C reservoir (Nelson and Carlson 2012; Powers et al. 2019). Sunken sargassum and its´ derived particulate organic matter also serve as a food source for deep sea fauna (Fleury and Drazen 2013).

The drifting sargassum rafts are often the only available natural surface substrate in open waters, enhancing the complexity of the pelagic environment. Besides from structural complexity, sargassum offers warmer microhabitats that provide advantageous for marine ectothermic organisms, including invertebrates and fish (Gulik et al. 2023). Consequently, the rafts play a crucial role in redistributing organisms within the pelagic environment, and may impact the survivorship of species dependent on it for both shelter and food. In the Gulf of Mexico and Sargasso Sea, 145 species of invertebrates and 80 species of fish have been reported in offshore (high-sea) rafts (Coston-Clements et al. 1991; Laffoley et al. 2011; Niermann 1986), including endemic and iconic species (Fine 1970; Hemphill 2005; Mansfield et al. 2021). The rafts are important recreational fishing areas for wahoo and jacks, and serve as a nursery and spawning area for a wide array of species, including commercially important ones (Laffoley et al. 2011; Wells and Rooker 2004).

Since 2011, a new sargassum concentration zone has formed in a region of the North Atlantic Ocean, near the equator, between Africa and Brazil, called NERR (North Equatorial Recirculation Region; Franks et al. 2016a, b; Gower et al. 2013). Since then, sargassum has been periodically introduced from the NERR into the Caribbean Sea, to be transported northwestwards by the prevailing currents, crossing the Caribbean Sea, into the Gulf of Mexico or the Gulf Stream area. This new area of sargassum concentration, from the NERR until the Gulf of Mexico, known as the Great Atlantic Sargassum Belt (GASB) contained more than 20 million tons of wet biomass during a peak month, in July 2018 (Wang et al. 2019). The environmental conditions in both the NERR and the Caribbean Sea differ from those in the Sargasso Sea and the Gulf of Mexico; the temperatures being generally higher in the GASB, which lies in the tropics, versus the sub-tropic waters of the Sargasso Sea, likely allowing for higher growth rates (Magaña-Gallegos et al. 2023). Nutrient availability varies throughout the concentration areas of sargassum, that receive nutrients from various sources, such as river plumes, upwellings and atmospheric depositions, that overall have been increasing in recent decades (Lapointe et al. 2021; Oviatt et al. 2019).

In this work, we characterize the associated fauna collected from rafts during two expeditions in the Caribbean Sea, in 2018 and 2019, two peak sargassum years in the GASB. The expeditions covered two ecoregions of the Caribbean Sea (Spalding et al. 2007) and samples were collected in the open-sea, as well as in coastal (< 20 km offshore) waters in one of them. Previous works have raised concerns about high concentrations of some toxic elements in beached sargassum in the Caribbean, especially arsenic (e.g., Rodríguez-Martínez et al. 2020; Alleyne et al. 2023a); therefore, we also analyzed the sargassum samples for metals and metalloids (metal(loids) from hereon) that are commonly associated with potential toxic effects (EPA 2024). There are various sources of nutrients in the NERR (Oviatt et al. 2019; Lapointe et al. 2021), but once entered into the Caribbean, where the depths of the thermocline and halocline are relatively stable, sargassum encounters more oligotrophic conditions than those in the NERR, although there are differences in nutrient availability within the Caribbean (Chollett et al. 2012). The hypothesis is that elemental contents (C, N, and possibly metal(loids)) and the community of small mobile fauna changes with ecoregion and distance from the shore within the Caribbean Sea.

The study of the associated faunal community of pelagic sargassum rafts and elemental tissue content of the algae throughout the Caribbean will increase our understanding of the biodiversity and processes in the open-sea and coastal sargassum rafts outside the better studied Sargasso Sea and Gulf of Mexico, and will provide further understanding on this relatively new proliferation of holopelagic Sargassum species.

Materials and methods

Description of sampling locations

Samples of sargassum and associated fauna were collected during daylight at two research scientific expeditions: 1. IMIPAS Expedition “Campaña América Central 2018” during September–October 2018, and 2. CEMIE-Expedition “CEMIE-Océano I” during June 2019. One nocturnal sample was taken during the CEMIE-Expedition, which was analyzed separately. Additional samples were collected exclusively for elemental analysis nearshore at Puerto Morelos, during March–May 2020 (Fig. 1, Supplement S1). During the CEMIE-Expedition, temperature, chlorophyll-a concentration ([chl]) and salinity were determined of the surface water with a CTD probe SN320-0913609 (Supplement S1); whereas for the IMIPAS-Expedition, the environmental data were obtained from the NASA Ocean Biology Processing Group database (Supplement S2). The area covered by all sampling stations was divided into two marine ecoregions (“Areas of relatively homogeneous species composition, clearly distinct from adjacent systems”) following Spalding et al. (2007): South Western Caribbean (SWC) and Western Caribbean (WC). In addition, the stations were classified whether they were in the open-sea (o: beyond continental platforms or > 20 km from the nearest shoreline) or coastal (c: above a continental platform, and < 20 km from the nearest shoreline) waters. Only the WC ecoregion had both open-sea and coastal stations. SWCo, WCc and WCo will be referred to as regions in this study.

Fig. 1
figure 1

Sampling locations in the Caribbean Sea during the INAPESCA-expedition (2018) and Cozumel Island during CEMIE-expedition (2019). See Supplement S1 for coordinates, and nearshore samples (PM-Puerto Morelos). WCc: Western Caribbean-coastal, WCo: Western Caribbean-open-sea, SWCo: South Western Caribbean-open-sea

Full size image

Sample collection

Sargassum with small mobile fauna at the IMIPAS-Expedition was collected with a 3.0-m long neuston net with a 1.0-m wide by 0.5 m-high opening. During the CEMIE-Expedition, a 3.0 m long neuston net of 1.5 × 1.0-opening was used. The mesh size of both nets was 3 mm. The nets were towed until a maximum of 10 kg of wet sargassum were collected (between ~ 1 to ~ 10 wet kg).

Holopelagic Sargassum species and morphotypes

The holopelagic Sargassum species and morphotypes, were identified on board, following the criteria described in Parr (1939) and Schell et al. (2015). At IMIPAS-Expedition, after identification, the samples were dried for posterior elemental analysis (subsections 2.3 and 2.4). At the CEMIE-expedition, the wet weight of each morphotype was also determined per sample, after separation of the small mobile fauna.

Trace metal(loid) concentrations in sargassum tissue

In the laboratory, sargassum samples were freeze-dried (3 days at constant weight at −45 °C and 70 × 10–3 millibar), ground to powder and homogenized. Aliquots of dried tissue (~ 300 mg) were acid digested in closed Savillex™ PFA vials, using 6 mL of HNO3 (Ultrapure > 65%, Trace metal Analysis, Baker™), for 12 h at room temperature and subsequently, for 3 h within a sand bath at 130 °C. The digestates were diluted with triple distilled water (~ 17.5 mL, to reach a final weight of 20 g) and stored in polyethylene bottles. Each batch of digestion included aliquots of certified reference materials (CRM, IAEA-359, IAEA-331, IAEA-413, DORM-4, DOLT-5) and one blank every 20 samples. All labware used during the analysis, was previously washed with a phosphate free detergent, rinsed with tap water, sequentially submerged for 72 h in acidic solutions (2 M HNO3 and 2 M HCl), and finally rinsed with triple distilled water.

The concentrations were determined through atomic absorption spectrophotometry (AAS); Cu and Zn by flame (Varian SpectrAA 220 (Clesceri et al. 1998); As, Se, Cd and Pb by graphite furnace (AAnalyst 800, PerkinElmer) (Bergés-Tiznado et al. 2015) and Hg by AAS by cold vapor generation (GVF, VARIAN model VGA-110).

The analysis of the 112 samples were performed in duplicate; when the duplicate results differed by > 8% (accounting for the systematic instrumental error), the analysis was repeated. The detection limit for each element was determined through replicated analysis (n = 10) of the lowest concentration used in the calibration curve (Supplement S3). The precision of the analysis was determined through the analysis of replicates (n = 10) and the calculation of the variation coefficient (CV (%) = mean/standard deviation × 100); and the accuracy of the analyses was evaluated through the recovery of each element in the certified reference materials (Supplement S4).

For each element, the readings below the limit of detection (< LD) were substituted with LD/√2 for calculation of summary statistics (Celo and Dabek-Zlotorzynska 2010). Multivariate techniques were employed to analyze variations in elemental composition. Relationships among metal contents in all sargassum samples spanning different regions were examined using non-metric multidimensional scaling (MDS) (Clarke et al. 2014), applying Euclidian distance measure. Subsequently, a one-way analysis of similarity (ANOSIM) was conducted to assess the statistical significance of observed disparities in metal contents across regions. Furthermore, a Similarity Percentage Analysis (SIMPER) was performed to identify the metals accountable for the observed differences in community composition. These analyses were performed with PRIMER v7.0.23 software (PRIMER-e).

C and N concentrations in sargassum tissue

Total and organic C and N concentrations were determined in aliquots of ~ 10 mg of dry and ground tissue, with an elemental analyzer Vario Micro Cube Elementar. The organic fractions ([Corg] and [Norg]) were determined in samples previously treated with 1 M HCl to eliminate the carbonate content, whereas the inorganic fractions ([Cinorg] and [Ninorg]) were estimated by the difference between total concentrations (([Ctotal] and [Ntotal]), and the organic fractions (Cuéllar-Martínez et al. 2019). The accuracy and precision of the total C and N analyses were evaluated through replicated analyses (n = 10) of the organic analytical standard (OAS) B2164 from Elemental Microanalysis Limited (Algae Bladderwrack, C = 33.67 ± 0.29%, N = 1.25 ± 0.02%). The concentrations obtained were within the certified values of both OAS, and the CV was 2.5%. Corg/Norg ratios were calculated based on their respective molecular weights.

One-way ANOVA was applied to test whether [Ctotal], [Corg] [Cinorg] and [Ntotal] in sargassum varied across species and regions; and whether [chl] in the water varied across regions. Correlations were done between [chl] in the water and the elemental (C, N) contents in sargassum that significantly differed among ecoregions in the ANOVA tests.

Fauna

The mobile fauna was extracted by placing the sargassum samples in trays or buckets with fresh water. The buckets were shaken vigorously to detach the mobile fauna from the fronds; the remainder of the sample was sieved through a 0.5-mm mesh to collect the remaining clinging fauna. Specimens were transferred in labeled bottles and fixed with 70%-ethanol. In the laboratory, the mobile macrofauna was identified with a stereoscopic microscope, using specialized keys: Castillo-Rodríguez (2014) for mollusks, De León-González et al. (2009) for polychaetes, Chace (1972), and Willliams (1984) for decapod crustaceans. The keys of LeCroy et al. (2000) for amphipods, Kensley and Schotte (1989) for isopods, and Froese and Pauly (2024) were used for fish. For each sample, the number of all organisms per taxon were determined. As the wet weights of sargassum were not determined during the IMIPAS-expedition, it was not possible to determine faunal density (number of individuals per kg of wet sargassum) for standardization purposes; therefore, their abundance was standardized to N = 1000 organisms per sample for further analysis.

Multivariate techniques were employed to analyze variations in macrofaunal assemblages. Relationships among communities spanning the regions were examined using non-metric multidimensional scaling (MDS) (Clarke and Warwick 2001). In these analyses, relative abundance data were fourth-root transformed, and the Bray–Curtis similarity measure was applied. Subsequently, a one-way analysis of similarity (ANOSIM) was conducted to assess the statistical significance of observed resemblance in macrofaunal assemblages across ecoregions within the Caribbean. Furthermore, a Similarity Percentage Analysis (SIMPER) was performed to identify the species accountable for the observed differences in community composition. These analyses were performed with PRIMER v7.0.23 software (PRIMER-e). Possible difference in the density of the taxa between the nocturnal sample and the diurnal one, taken at approximately the same position, was analyzed with the function Wilcoxon signed-rank test, using the RStudio statistical software.

Results

Sargassum species and morphotypes

During both expeditions, IMIPAS (September–October 2018) and CEMIE (April 2019), the three morphotypes Sargassum fluitans III, S. natans I, and S. natans VIII were found at almost all stations, although at some stations S. natans VIII was absent. During the CEMIE-Expedition, S. natans I (X ± SE = 46.3 ± 3.7%, n = 12) and S. fluitans III (41.1 ± 3.6%) were almost equally represented, and S. natans VIII was the least abundant morphotype (12.6 ± 2.5%), although there was considerable variation in morphotype representation among the samples (Fig. 2).

Fig. 2
figure 2

Proportional biomass of the holopelagic Sargassum species and morphotypes at the stations of the CEMIE-1 expedition in April 2019. 21N was a nocturnal sample at station 21

Full size image

Metal(loid) concentrations

The seven analyzed elements were detected in the most of the sargassum tissue samples. Four samples of S. natans I (sites S26, S27, S28, and S29) for Se; and seven samples for Hg of S. natans VIII (sites S13, M34) and of S. natans I (sites S10, S12, S18, S19, M35) were below the limit of detection. The element concentrations obtained for the sargassum samples including all stations and all morphotypes generally followed the descendent order As > Zn > Cu > Cd > Se > Pb > Hg (Table 1). Total concentrations of As were high across species and regions with a mean concentration of 162.4 µg g−1 (SE = 5.7, n = 107). ANOSIM revealed that the elemental composition varied significantly among morphotypes (overall R = 0.221, P = 0.001; S. fluitans III vs S. natans 1, R = 0.327, P = 0.001; S. fluitans III vs S. natans VIII, R = 0.148, P = 0.002; S. natans I vs S. natans VIII, R = 0.166, P = 0.005). Differences in trace element composition also differed significantly among regions (overall R = 0.218, P = 0.001; SWCo vs WCo, R = 0.199, P = 0.001; SWCo vs WCc, R = 0.183, P = 0.001; WCo vs WCc, R = 0.305, P = 0.001) (Table 1). The SIMPER analysis revealed that Zn, Pb and Hg differed most amongst the sargassum morphotypes, explaining together 89.7% of the variation, and Hg accounted for 67.7%. The element that most explained the differences amongst the regions was Hg, explaining 92% of the variation among the regions, in the order SWCo > WCo > WCc (Table 1).

Table 1 Mean (± SE) concentrations of metals and metalloids in sargassum tissues (µg g−1). Sf3 Sargassum fluitans III, Sn1 S.natans I, Sn8 S. natans VIII, SWCo SouthWestern Caribbean-open-sea, WCo Western Caribbean-open-sea, WCc Western Caribbean-coastal. As is highlighted because its high concentrations in sargassum have been of concern for potential valorization applications (Derochers et al. 2021). N number of samples
Full size table

Concentration of C and N in sargassum and of chlorophyll-a in the water

Overall mean content of Corg in sargassum was 19.32 mg g−1 (SE = 0.31, n = 102), of Cinorg 8.58 ± 0.26 mg g−1, Norg 0.824 ± 0.019 mg g−1, Ninorg 0.36 ± 0.01 mg g−1, and mean Corg/Norg was 29.19 ± 0.92. Two-way ANOVA did not show significant differences in [Corg] among morphotypes and regions, but [Cinorg] varied among morphotypes (F(2,93) = 29.388, P < 0.001) and regions (F(2,93) = 25.493, P < 0.001), with a significant interaction between morphotype and region (F(4,93) = 5.753, P < 0.001). Mean [Cinorg] was 8.58 ± 0.33 mg g−1 for S. fluitans III, 7.42 ± 0.33 mg g−1 for S. natans I, and 10.72 ± 0.53 mg g−1 for S. natans VIII; all pairwise comparisons were significantly different at P = 0.05 (Tukey test). [Ntotal] only varied significantly with region (F(2,93) = 5.837, P = 0.004). Chlorophyll concentrations in the water varied across regions (F(2,99) = 13.438, P < 0.001), and there was a general tendency of increasing [Cinorg] and [Ntotal] with increasing [chl] in the order SWCo > WCo > WCc (Table 2). However, only the correlation between [Cinorg] and [chl] was significant (Pearson correlation = 0.258, n = 102, P = 0.009).

Table 2 Mean ± SE concentrations (mg g−1) of total carbon [Ctotal], total nitrogen [Ntotal], inorganic carbon [Cinorg], and molar total C: total N ratio (C:N), in sargassum tissues, and chlorophyll-a concentration [Chl] (mg m−3) in superficial sea water per region. SWCo SouthWestern Caribbean-open-sea, WCo Western Caribbean-open-sea, WCc Western Caribbean-coastal
Full size table

Fauna

In total, 39,134 mobile macrofaunal organisms were collected from the sargassum rafts during both expeditions, comprising 66 taxa belonging to 11 phyla (Table 3, Supplement S5). The phylum Arthropoda stood out for its taxonomic richness and numerical dominance, with 32 taxa (29 of them belonging to the subphylum Crustacea), constituting 60.8% of the collected fauna. This was followed by Mollusca with 21.5%, Chordata with 10.4%, and Annelida with 6.0%. The remaining phyla accounted for less than 0.1%. Notably, the Sargassum Shrimp Latreutes fucorum (33.3%) and the Brown Sargassum Snail Litiopa melanostoma (21.2%) were the most dominant species. Other species, such as the isopod Carpías minutus, constituted 13.5% of the total abundance, yet exhibited lower abundance in the SWCo region (Fig. 3). A total of 15 ichthyofauna (phylum Chordata) taxa were documented, representing eight families (Antennariidae, Carangidae, Coryphaenidae, Clupeidae, Exocoetidae, Hemiramphidae, Tetradontidae, and Syngnathidae). The majority of individuals (> 90%) were in juvenile stages. The Planehead Filefish (Stephanolepis hispidus) was the dominant species, comprising 33.3% of the recorded Chordata abundance, followed closely by Caranx bartholomaei (14.2%) and C. crysos (12.5%).

Table 3 Mean standardized abundance (to a total of 1000 individuals) and occurrence (% of stations present) of the sampled associated species with sargassum rafts in the Caribbean Sea during the during the INAPESCA-Expedition, September–October 2018 and CEMIE-Expedition in April 2019. SWCo: South Western Caribbean-open-sea, WCo: Western Caribbean-open-sea, WCc: Western Caribbean-coastal (excluding the nocturnal sample). N (in bold) indicates the total number of sampled organisms per region in the columns of Standardized abundance. In the columns of Occurrence, N indicates the number of sampled stations per region
Full size table
Fig. 3
figure 3

Shade plot illustrating the distribution of most abundant taxa (accounting for ≥ 1% of the total abundance) across the three regions. White space denotes absence of that species at that station; intensity of the grey scale is proportional to a fourth-root transformation of relative abundance. WCc Western Caribbean-coastal, WCo Western Caribbean-open-sea, SWCo South Western Caribbean-open-sea

Full size image

By region, taxa richness was highest in WCc (S = 42 taxa), followed by SWCo (S = 32) and WCo (S = 30). Eighteen out of the 66 taxa (27.2%) were common across all three regions (Table 3). WCc had a higher taxonomic overlap with SWCo (N = 21) compared to WCo (N = 19). Furthermore, WCc documented the highest number of unique taxa (18).

The nMDS ordination analysis separated stations from WCc region, and showed overlap in the mobile macrofauna composition between regions WCo and SWCo (Supplement S6). This was confirmed by the ANOSIM revealing overall significant differences in sargassum-associated mobile fauna composition in the three regions (R = 0.322, P = 0.001). Posterior pairwise tests discerned differences between the coastal zone (WCc) and both open-sea zones (WCc vs SWCo, R = 0.424, p = 0.001; WCc vs WCo, R = 0.491, P = 0.001), but not so between the rafts in both open-sea regions (SWCo vs WCo, R = 0.063, P = 0.172; Table 1). The SIMPER analysis revealed that the Sargassum Shrimp Latreutes fucorum was the primary contributor to dissimilarity between the coastal (WCc) and open-sea (WCo) western Caribbean regions, displaying greater relative abundance in the WCc region in contrast to WCo (Table 4). Litiopa melanostoma, Carpias minutus, some fish species and copepods also were discretionally present in the studied Caribbean regions (Fig. 3, Supplement S5). The other two (almost) omnipresent species, polychaeta Platynereis dumerilli and the Planehead Filefish Stephanolepis hispidus, were equally abundant across the three regions (Fig. 3).

Table 4 The contribution of sargassum-associated macrofauna species to the observed similarity among the three regions in the Caribbean, calculated by Similarity Percentages (SIMPER)
Full size table

The nocturnal sample at Station 21 (CEMIE-Expedition), yielded 32 taxa, most of which were also reported during daytime, with dominance of Carpias minutus and L. fucorum. However, the nocturnal sample also included decapod crustaceans in larval stages (zoea and megalopa) from various families (Penaeoidea, Paguridae, Porcellanidae, Xanthidae), together with a lobster phyllosoma larva (Panulirus sp.) and stomatopoda (Squillidae) larva, which were absent in diurnal samples. At station 21, the nocturnal sample was more diverse (32 taxa) than the diurnal sample with 19 taxa, but the density was lower (184 individuals per wet kilogram of sargassum) than the daylight sample (304 individuals per wet kilogram of sargassum; Wilcoxon U = 188.5, N = 32, P = 0.0244; Supplement S5).

Discussion

General

We characterized biochemistry and associated fauna of sargassum during two consecutive high influx years (2018 and 2019; Rodriguez-Martinez et al. 2022; Wang et al. 2019), during two expeditions throughout the Caribbean, and during both expeditions extensive rafts were encountered. The three most common morphotypes in the GASB, were encountered in most samples, although S. natans VIII was absent in several samples of the IMIPAS-expedition, and had lowest relative abundance in the samples of the CEMIE-expedition. S. natans VIII, a rare morphotype in the Sargasso Sea, dominated the first years of the sargassum bloom in the GASB (Schell et al. 2015). However, for unknown reasons, this morphotype slowly has become less dominant in the following years (Garcia-Sánchez et al. 2020). The varying composition of Sargassum species and morphotypes may be associated with environmental conditions, because the morphotypes have different physiology, reflected, amongst others, by different growth rates under varying conditions (e.g., Magaña-Gallegos et al. 2023), and elemental contents (Cipolloni et al 2022; Rodriguez-Martínez et al. 2020; this study). The sargassum composition may also be relevant for the associated fauna, because some epibionts have specific affinity for a morphotype (Mendoza-Becerril et al. 2020). Martin et al. (2021) reported higher motile fauna abundance and species richness in rafts composed of mostly S. fluitans III, than with mixed rafts having more equal representation of two or three morphotypes (but see Van Tussenbroek et al. 2024).

Metals and metalloids

The metal(loid) concentrations varied significantly among morphotypes, and among regions. The metal sequestration involves ion exchange, chelation, adsorption, and entrapment that depend on properties of cell wall constituents (Davis et al. 2003), which vary among species and environments. Zn, Pb and Hg differed most among the morphotypes, and high variability in metal(loid) contents among morphotypes has also been reported in other studies (e.g., Cipolloni et al. 2022; Rodríguez-Martínez et al. 2020). In the present study, the mean concentration of As ranged in the different ecoregions and morphotypes from 139.6–159.6 µg g−1 for the S. natans I, 152.7–195.5 µg g−1 for the S. fluitans III, and 159.8–186.8 µg g−1 for the S. natans VIII in all regions. These concentrations were usually within range of those recently observed in stranded sargassum in Mexico (Ortega-Flores et al. 2022; Rodriguez-Martinez et al. 2020), Barbados (Alleyne et al. 2023a), in rafts in the NERR (Cipolloni et al. 2022; Dassié et al. 2022; McGillicuddy et al. 2023), southeastern Caribbean (McGillicuddy et al. 2023), and Sargasso Sea (McGillicuddy et al. 2023). McGillicuddy et al. (2023) suggested a relationship between high arsenic uptake and low phosphate availability, typical for oligotrophic Caribbean waters. High arsenic is of special concern, as it may accumulate in the environment (Hernandez-Almaraz et al. 2016). The levels in this study are above the limits recommended for agricultural soils in some countries (Cipolloni et al. 2022; Rodríguez-Martínez et al. 2020). This finding confirms that caution is needed when using the Caribbean sargassum for use as human or animal food source/additives (Derochers et al. 2021), although As can be removed by simple procedures from the algae (Cisneros-Ramos et al. 2024).

Fauna

The mobile macrofauna associated with open-sea rafts in the Caribbean encompassed a diverse array of organisms, most of them commonly associated with oceanic or open-sea rafts. In this study, small copepods (Calanoida) and ostracods (Halocyprida) were occasionally collected, but the mesh size was too large for systematic collection of these organisms; thus, it was not possible to establish whether these are typically associated with sargassum rafts. Although many species of calanoid copepods have planktonic lifestyles, they can be found incidentally in floating mats of algae (Thiel & Gutow 2005). Some harpacticoid species are typically associated with sargassum, such as Scutellidium sargassi and Macrochiron sargassi, which has prehensile maxillipeds terminated by a long claw, allowing for grasping the algae (Yeatman 1962). Other incidental invertebrate visitors were siphonophores (Abylidae, Diphyidae), amphipods (Hyperiida), decapods (Belzebub faxoni, Lucifer typus), insects (Halobates micans), and mollusks (Pteropoda), which are not typically associated with sargassum but are commonly found in the pelagic environment (Felder et al. 2009; Flores-Coto et al. 2013; Gasca 2003; Stoner and Humphris 1985).

The density of fauna in the rafts was lower in the night than the day sample taken at approximately the same position. Although only a single night sample was collected, so caution is needed to draw any firm conclusions. The detected decreased density may be due to the fauna leaving the sargassum rafts to feed during the night, as do benthonic organisms leaving their bottom shelter overnight (Gibson et al. 2016; Queiroga and Blanton 2005). The nocturnal sample contained 11 unique taxa which were primarily meroplanktonic larvae, with their adult counterparts predominantly associated with benthic habitats, suggesting a to-and-fro nocturnal migration of benthic and pelagic organisms to the sargassum rafts.

The sampled ichthyofaunal diversity was lower than that reported in the Gulf of Mexico, which aligns with the findings of Alleyne et al. (2023b) and Van Tussenbroek et al. (2024), who studied the free-swimming fish that accompanied the near-shore rafts (within several kms from shore) with underwater cameras in Barbados and Mexico, respectively. Alleyne et al. (2023b) suggested that the low diversity of fish could be attributed to the recent origin of rafts in the GASB region. More fish were observed in the open-sea as compared to the coastal samples, which agrees with Wells and Rooker (2004) who observed a significantly higher abundance of fish associated with sargassum further offshore, although Lapointe et al. (2014) found higher fish abundance in neritic waters. In this study, mostly juvenile fish were sampled, which aligns with the findings of other studies in the Sargasso Sea and Gulf of Mexico (Bortone et al. 1977; Casazza and Ross 2008; Wells and Rooker 2004). Lapointe et al. (2014) also reported high numbers of juvenile filefish Stephanolepsis hispidus (Monacanthidae) and jacks (Carangidae) associated with sargassum, and they found that the fish excretions were an important source of N and P for sargassum, sustaining sargassum growth. Various crustaceans/invertebrates were also mainly present in juvenile stage, such as Platynereis dumerilii, Sunampithoe pelagica and Portunus sayi. In this study, occasional loose eggs as well as egg sacs were observed. This, together with presence of many juvenile specimens, confirms the function of the Caribbean sargassum rafts as a significant nursery habitat, as well as spawning area for fish (Oxenford et al. 2019) and other organisms.

Patterns in biogeography within the Caribbean

Differences in elemental composition and associated mobile fauna were encountered among regions within the Caribbean. The ecoregions corresponded mostly with Caribbean open-sea provinces based on general physicochemical properties (water transparency, salinity, wind-driven wave-exposure and hurricane incidence) by Chollett et al. (2012; Supplement S1). Major sources of heterogeneity include river plumes (Hu et al. 2004; Restrepo et al. 2006), eddy development (Andrade and Barton 2000), and upwelling (Müller-Karger et al. 1989; Andrade and Barton 2005). The Western Caribbean (WC; Physiochemical Province 3 of Cholett et al. 2012) tends to have slightly higher temperatures than the South Western Caribbean (SWC), as well as clearer waters, and is likely more oligotrophic. Wind driven wave exposure and salinity is similar for both physiochemical provinces, except for southernmost stations (S27-29) presenting slightly lower salinities. However, most differences within the physiochemical provinces are fuzzy, with smooth transitions.

The absence of a consistent tendency in metal(loid) contaminants in sargassum samples confirms such fuzzy regional transitions, and has been observed by previous studies, linking this to the pelagic nature of sargassum rafts and their trajectories (e.g., Rodríguez-Martínez et al. 2020; Cipolloni et al. 2022). Even though some differences were found among ecoregions, especially in Hg content, which might be the result of variability in anthropogenic (e.g. waste treatment outfalls or industrial point discharges) and natural (geological setting, mineralogical composition of rocks and soils, and the fluvial inputs) sources. Natural high Hg levels are commonly associated with plate tectonic boundaries, volcanic activity, mineralized areas, or soils enriched with Fe-Al-oxyhydroxides or pyrite (Fitzgerald and Lamborg 2014; Gustin et al. 2000), and dry (dust) deposition (Zhang et al. 2023). The Western Caribbean is further away from such sources than the Southwestern Caribbean. This, in combination with growth dilution (decreasing content due to increasing weight through growth), may explain the differences between the regions, where S. fluitans III collected from the South-Western Caribbean-open-sea region (SWo) exhibited the higher Hg levels (0.22 ± 0.09 µg g−1), than the Western Caribbean-open-sea (0.13 ± 0.02 µg g−1), and also the Western Caribbean-coastal region (0.06 ± 0.01 µg g−1).

The surface water chl a concentrations throughout the Caribbean Sea during the time of this study (0.06–0.12 mg m−3), were lower than those generally reported in the NERR (0.14–0.30 mg m−3; Jouanno et al. 2021), but higher than that reported in the southern Sargasso Sea (0.046 ± 0.004SE; Morel et al. 2010). In the Caribbean Sea, the chl a concentrations were lower in the western than south western Caribbean (SWC), and sargassum total N content followed the same tendency, indicating regional differences in nutrient availability. Cinorg content in the sargassum samples tended to be higher in regions with higher chl a concentration in the water, which may indicate higher abundance of calcifying epibionts such as, serpulids, bryozoans or hydroids (Pestana 1985). These epibionts may reach higher abundance in more nutrient rich waters, although not much is known about causing factors of epibiont abundance and composition, which can be highly variable (e.g., Huffard et al. 2014).

The associated mobile fauna community was similar for both Caribbean open-sea ecoregions, but differed from that collected above the continental platform in the Western Caribbean. The samples above the continental platform in the Western Caribbean during the CEMIE expedition were collected in the following year of the IMIPAS-Expedition, and by a different crew, but following the same collecting protocol; thus, some differences in fauna communities between the two expeditions are possibly explained by different sampling time or on-board sample processing. This may specially apply to the associated fauna. However, during both expeditions, sample size was large (> 1000 specimens per sample; Supplement S5) therefore, they were all representative. The specific richness in the rafts was higher above the continental platform (WCc) than in the open-sea rafts, which may be explained by the colonization by benthic fauna from the continental platform (Queiroga and Blanton 2005), as early benthic stages of many decapods commonly settle on floating substrates (Thiel and Gutow 2005; Wehrtmann and Dittel 1990). For example, Franks and Flowers (2008) discovered Cerataspis monstrosa larvae of the deep-sea shrimp (Plesiopenaeus armatus) in the gut contents of dolphinfish (Coryphaena spp.) known to feed in sargassum rafts (Morgan et al. 1985).

Ecological importance of open-sea sargassum rafts in the GASB

Since the initiation of the proliferation of sargassum in the GASB, the coasts of the Caribbean (and also those of Brazil and Africa) have suffered, on a recurring basis, from massive beaching and decomposition of these algae, with negative consequences for the coastal ecosystems, society, tourism industry, and human health (Van Tussenbroek et al. 2017; Resiere et al. 2018; Chávez et al. 2020; Devault et al. 2021; UNEP-CEP 2021). The region-wide consequences of the massive beaching have spurred various research efforts into the study of sargassum, mainly focused on its negative impact on the coastal life and ecosystems, human health, and economies, following and predicting the blooms through satellite monitoring and modelling, and potential valorization of the algae (summarized in UNEP-CEP 2021). The unexpected bloom and its negative impacts have led to a disregard for the potential ecological importance of the open-sea rafts (but see Alleyne et al. 2023b; Martin et al. 2021).

The open-sea sargassum in the GASB may serve as a carbon sink, thereby mitigating global climate change (Gouvêa et al. 2020), although its importance for global carbon sequestration is debated (Bach et al. 2021; Marsh et al. 2023; Hu et al. 2021). Even though its role in global C sequestration may not be very significant, it is beyond doubt that sargassum in the GASB attains significant biomass (Wang et al. 2019) and high growth rates (e.g. Corbin and Oxenford 2023; Magaña-Gallegos et al. 2023), with consequential high productivity of particulate organic matter and dissolved organic carbon, that is expected to be exported to the deep sea or to enter the food web, as has been reported for the Sargasso Sea and Gulf of Mexico (Baker et al. 2018; Powers et al. 2019; Wells et al. 2017). In the Caribbean, the open-sea rafts had a diverse mobile fauna, similar to that in the Sargasso Sea and Gulf of Mexico.

In the Sargasso Sea, the pelagic rafts are viewed as environmentally, socially, economically valuable ecosystems that should be preserved and managed (Acton et al. 2019; Freestone 2021; Laffoley et al. 2011; Pendelton et al. 2014; Trott et al. 2011). The full scope of ecosystem services of open-sea rafts in the GASB have yet to be assessed, but they likely parallel those of the rafts in the Sargasso Sea. For the Caribbean in particular, of which most waters are assigned EEZs (Exclusive Economic Zones, van der Plank et al. 2022), recognition of the ecological importance of the open-sea rafts may be a first step in accomplishing international agreements and clarity of rules on the management of sargassum. In the USA, in-water sargassum harvesting is forbidden by NOAA; however, in the Caribbean, collection at sea, in addition to beach collection, is necessary to mitigate ecological, social and economic disastrous consequences of the sargassum inundations. Since 2011, much effort has been put in following and predicting the coastal inundations (Marsh et al. 2022 for review; Lara-Hernandez et al. 2024). Such models can also predict (with a certain probability) which rafts will likely never reach the shore. Following a precautionary approach to governance, sargassum in the open sea or ocean in the GASB, destined to remain drifting in the ocean, merits similar protective efforts as the oceanic rafts in the Sargasso Sea.